ALLELE SPECIFIC DETERMINANTS OF HOMOTHALLISM IN SACCHAROMYCES LACTISI

ALBERTA HERMAN293 AND HERSCHEL ROMAN

Department of Genetics, University of Washington, Seattle Received November 29, 1965

N general, fertile yeasts are classified as being either heterothallic or homo- thallic. Heterothallic strains exhibit one of two alternate mating types (usually

’ a! designated as a and or + and -) with diploidization occurring between cells of opposite mating type only (e.g. a with a! and + with -) . In homothallic yeasts, on the other hand, restrictive mating systems are not observed and sister cells quite commonly fuse to form zygotes. The known genetic controls regulating homothallism in yeast are numerous and apparently complex in their action. For example, LEUPOLD(1958) working with the fission yeast Schizosaccharomyces pombe strain liquefaciens found that mating behaviour was controlled by a complex locus. Two heterothallic strains, h+ and h-, were genetically determined by closely linked mutant sites within this complex locus. Homothallic strains, designated H40 and Hgo,could be derived from these heterothallic cultures either through mutation or through recombina- tion within the complex region. In the yeast Saccharomyces chevalieri the pres- ence of a gene D in a single-spore culture committed the culture to homothallism (WINGEand ROBERTS1949). The D gene was independent of the mating locus and was equally effective in the presence of either the a or alleles. Heterothallic cells of S. chevalieri, on the other hand, possessed the recessive d allele and exhibited either one of the alternate mating types a or a. OESER(1962) and HAWTHORNE(1963b) have reported that the D gene produces homothallism by directing mutation of a mating type allele to the alternate form in some of the cells within a culture. Following this mutation, the affected culture becomes homothallic since it consists of a mixture of a and a! sister cells capable of zygote formation. In Saccharomyces cerevisiae homothallism results either from a high mutation rate of one mating type allele to the other (AHMAD1952) or from the action of the complementary genes HM,HM, and HM,HM, (TAKAHASHI,SAITO, and IKEDA1958; TAKAHASHI1958). To compare the nature of the D and HM genes, TAKAHASIand IKEDA( 1959) prepared hybrid Saccharomyces diploids containing both the HM,HM, complement and the D gene of WINGEand ROBERTS(1949). The D gene and the HM genes segregated independently. This paper reports the identification of factors affecting homothallism in

’ This recearch supported by Public Health Service grants 1459 and 5R01 (AT-00328-14) Public Health Service postdoctoral fellowship 5-FZ-GM-13, 112-02 Present address Northern Utilization Reseaich and Development Di\ision, ‘IRS, U S Department of Agriculture, Peoria, Illinois

Genetlcs 53: i27-740 Apnl 1966 728 A. HERMAN AND H. ROMAN selected strains of the yeast Saccharomyces Zactis. Two unlinked independent homothallic genes were identified. These factors exert their effect by changing the mating type to the alternative form in a certain percentage of the cells. Moreover, these two genes are allele specific. One locus is effective in changing the mating

type from a to (Y, the other in changing a to a.

MATERIALS AND METHODS

Parental stocks Y14 and Y123 were obtained from L. J. WICKERHAM(strains Y1140 and Y1118 respectively). All isolates used in this study were derived from matings in which these two strains were the original parents. The maintenance of stocks and the procedures used in tetrad dissections and carbohydrase assays were described previously (HERMANand HALVORSON 1963). Cells grown for 24 hours on YM (WICKERHAM1951) were used as mating inocula. Malt extract agar (ME) (WICXWMAN1%1) was used for mating and sporulation. Vegetative growth in Saccharomyces lactis consists essentially of haploid cells. The diploid phase is limited, in the majority of cases, to the zygote formed by the mating of two haploid cells. Usually the zygote forms four haploid ascospores per ascus. Since mating is immediately followed by sporulation, ME serves both as a mating and sporulating medium. In both homothallic cultures and hetero- thallic mating mixtures, zygote formation was confined exclusively to cells grown on ME. Cells inoculated onto ME do not immediately mate but undergo a period of multiplication (8 to 10 hr) before the first competent cells appear in the population. Whether the cells from the original inoculum ever become competent or whether the mating response is confined exclusively to cells formed during growth on ME is not known. Following the 8 to 10 hr period of vegetative growth the proportion of cells capable of producing zygotes gradually increases until a maximum degree of competency is attained after 3 to 5 days incubation. Homothallic and heterothallic cultures are differentiated on the basis of differences in their mating response during growth on ME at 28°C. Homothallic cells form zygotes when grown alone on ME. Heterothallic strains exhibit no mating response under these same conditions but produce zygotes after inoculation in mixed culture with one of the two tester strains, either a or a. As will be seen later, most of the homothallic strains were weakly homothallic, i.e., less than 25% of the cells in a population formed zygotes. During this study it was necessary to determine the homothallic potentialities of several segregants. This was done by counting the number of zygotes formed per 100 vegetative cells after 3 to 5 days incubation at 28°C and is referred to as the percent homothallism of a culture. In addition to determining the percent homothallism, several cultures were also examined for the presence of cells possessing a or a mating potentialities. For this determination, standard inocula of the strain to be tested were mixed and plated with equal volumes of the a or a testers on ME. After 4 days incubation at Doc,the ratio of zygotes to vegetative cells was determined in a sample of each cell mixture and compared with the ratio in the homothallic strain grown by itself. The rationale of this procedure was as follows: In the presence of an excess of a tester cells, each cell in the homothallic population possessing CY characteristics should mate. Similarly, a challenge with an excess of a tester cells should initiate zygote formation by each cell with a qualities. Thus any marked increase in the zygote : vegetative-cell ratio in mixed matings com- pared to homothallic control implied the presence of cells possessing a or a mating potentialities in the homothallic culture. Usually 400 to 500 cells were counted per sample. Larger numbers were counted when the ratio of zygotes to vegetative cells was less than 1% in the homothallic strain. Generally the homothallic segregants contained an excess of competent a or CY cells. Those possessing more a cells were designated as a-associated homothallic strains; those with more a cells as @-associatedhomothallic cultures. Tetrads exhibiting normal segregation patterns for adenine, histidine, leucine and/or trypto- phan nutritional markers as well as for the carbohydrases described previously (HERNIANand HALVORSON1963) were used in this study. With the exception of one adenine requiring mutation, none of the markers used was closely linked to the mating locus. The adenine mutant controlled HOMOTHALLISM IN YEAST 729 the production of a red pigment and was linked to the mating locus with a recombination fre- quency of about 4%. Each stock is identified by a tetrad number and a spore-culture letter A through D. The crosses and tetrads derived from each hybridization which were used in this study are listed in Table 1. In all the hybridizations considered in this paper, the complete genotype relating to mating type is listed. This is followed, in parentheses, by the stock number of the respective parents used in each cross.

RESULTS The parental mating types: Parental cultures Y14 and Y123 mated and sporu- lated when they were mixed and grown on ME plates. Neither strain showed a mating response when grown alone on ME. They were, therefore, identified as heterothallic cultures. Strain Y14 was designated as the a mating type, strain Y123 as the a: mating type. Evidence for homothallism: When the mating type of progeny derived from the parental cross a X a: (Y14 x Y123) was determined, both homothallic and heterothallic segregants were identified. Segregation ratios for homothallism: heterothallism included 3:l (3 tetrads), 2:2 (50 tetrads), 1:3 (77 tetrads), and 0:4 (26 tetrads). The 3 homothallic: 1 heterothallic ratio indicated that the hybrid was heterozygous for more than one locus for homothallism while the 0 homo- thallic:4 heterothallic ratio implied that the homothallic condition required spe- cific combinations of the homothallic factors and the mating type alleles. In the 26 tetrads in which all of the progeny were heterothallic, segregation of a and a: mating types could be followed readily. The ratio of a:a: was always 2:2 and indicated that a and a: were segregating as alleles of the mating locus. It was inferred that a and N segregated in this manner in all of the tetrads even though

TABLE 1 Hybridiurtions and tetrads used in this investigation

___ Hybridization Tetrad number ah,H, x aH,h, (Y14 x Y123) 155 157 31 1 338 599 729 1009 cuh,H, (729B*) 1216 aH,h, (155C') 25 1 &,ha (338C') 95 3 cuH,h, (251A*) 1254 1256 aH,h, (1256C') 1290 aH,h, X ah,H, (251A' x Y14) 1255 1257 1258

* IIomothallic. 730 A. HERMAN AND H. ROMAN this segregation was obscured in many tetrads by the presence of the homothallic condition. TOaccount for the above observations, the following hypothesis is proposed: In addition to the mating type alleles a and a, there are two nonallelic genes for homothallism, Ha and Ha. The combinations CUH, (either ahaH, or aHaHa) and da(either &aha or aHaHa)result in homothallism. Based on this hypothesis, the genotype of parent Y14 is ahaHa and Y123 is aHah, and both are therefore heterothallic. On this interpretation, the cross aha, x allaha (Y14 x Y123) would be ex- pected to yield haploid progeny of the following genotypes: (1 ) ahaH,, (2) ah,&, (3) aHaha, (4) ahah,, (5) aHaH,, (6) aHah,, (7) (JI,H,, and (8) ahaHa. The first four would be heterothallic and the last four homothallic segregants. All possible tetrad combinations and the frequency with which these tetrads were obtained are given in Table 2. The linkage relations between the three genes can be determined from the number of parental ditype (PD), nonparental ditype (NPD) and tetratype (T) segregations shown in Table 2. The PD:NPD:T ratio for a in relation to h, is 28:31:97; it is 129:0:27 for a in relation to h,; and 48:28:80 for Ha in relation to Ha. Linkage between two genes is demonstrated when the parental ditypes are significantly more frequent than the nonparental ditypes (PERKINS1953). It can be seen that a and ha exhibit no linkage, whereas a and h, are strongly linked (27/156 x i/z = 9%). There is also some indication of linkage between Ha and H, but at best these are loosely linked.

TABLE 2 Type of segregations observed in the mating ah,H, x aH,h, (Yf4 X Y123)

Type of tetrad

~~ ~~ ~ Numbers observed 0 0 3 0 20 22 2 75 26 Ratio of homothallism: heterothallism 4:O 3:l 3:l 2:2 2:2 2:2 1:3 1:3 0:4 Ratio of homothal1ism:a:a 4:O:O 3:l:O 3:O:l 2:2:0 2:0:2 2:l:l 1:2:1 1:1:2 0:2:2

PD, NPD or T’ relationship for genes a:ha NPD NPD NPD PD NPD T PD T PD a:H, NPD NPD T NPD PD T T PD PD ha:H, PD T T NPD NPD PD T T PD

* PDzparental ditype; NPD=nonparental ditype; T=tetratype. HOMOTHALLISM IN YEAST 73 1 If all three genes are on the same chromosome, the probable order is a-Ha-h,. With these linkage relations in mind, it is easy to understand why some of the classes in Table 2 'were not obtained. These would have required a four-strand double crossover between the mating-type locus and the H, locus; such events would be expected to be rare. The other minority classes require a single cross- over in this region. Independent segregation of Ha and Ha: To test the foregoing hypothesis that there are two nonallelic genes governing homothallism, tetrads 157 and 1009 derived from the cross Y14 x Y123 were examined for the genotypes of some of their segregants. In the first tetrad, 157, the mating segregation was 2 homothallic: 2a (Table 3). We can infer from this segregation that 157A and 157D, which are homothallic, each contained an a allele and H,. Spores 157B and 157C, the heterothallic segre- gants, should therefore each possess an h, allele and, since they are of a mating type, an h, allele as well. The proposed genotypes of the four segregants are summarized in Table 3. Neither 157B nor 157C should produce a-associated homothallic progeny when mated with heterothallic ah, cells. Eighteen tetrads from the cross ahaha x ah,H, (157B x Y14) and 15 asci from the cross ahaha x ah,H, (157C x Y14) were dissected and the mating types of the segregants determined. The complete segregations for mating type observed in these crosses were as follows: The mating ahah, X ah,Ha (157B x Y14) pro- duced 2a:2a (16 tetrads) and 2a: la: 1H (2 tetrads) ratios, while from the cross ahah, X ah,Hol (157C X Y14) %:2a (13 tetrads) and 2a:la:lH (2 tetrads) ratios were obtained. As predicted, none of the a-bearing segregants were homothallic. In the second tetrad, 1009, the mating types of the segregants were 2a:la:lH. The proposed genotypes of the four spore-cultures are given in Table 3. Both 1009A and 1009D carried the h, allele since they were heterothallic and of a mating type. One of these segregants should also contain an H, gene. To deter- mine which segregant contained H,, each strain was mated with ah,H, (Y14) cells and their progeny examined for a-associated homothallism. The segregations from the zygote &,ha x ah,H, (1009A x Y14) yielded 2a:2a (6 tetrads), la:2,: 1H ( 18 tetrads) and 2a:2H (8 tetrads) ratios while the cross

TABLE 3 Mating type of spore isolates from tetrads 157 and 1009

-___ Tetrad Mating type Proposed genotype 157A H' aH,Ha 157B a ah&, 157C a ahah, 157D H aHnHa 1009A a aH,ha 1009B H aH,Ha 1009c a ah,H, 10WD a ah,h,

' €I = homothallic. 732 A. HERMAN AND H. ROMAN ah,h, X ahH, (1009D X Y14) gave 2a:2a (31 tetrads) and 2a:la:lH (4tetrads) ratios. Twenty-six of the 32 tetrads derived from the mating of 1009A x Y14 thus contained a-associated homothallic segregants while none of the 35 tetrads obtained from the cross 1009D x Y14 possessed a-associated homothallic isolates. As predicted, only one of the two strains, 1009A, carried the Ha allele. Therefore the complete genotype of 1009A was aH,h, while that of 1009D was ahah,. A further test of the genotypes of the 1009 tetrad was made by hybridizing 1009A X 1009C. The inferred genotypes are aHah, and ahH, respectively. A mating of these two heterothallic cultures should be equivalent to that of Y123 X Y 14 and should produce both a-associated and a-associated homothallic recombi- nants. This proved to be the case. Forty-five tetrads from the mating 1009A x 1009C were dissected and the mating types of the spore isolates determined. The segregations obtained from this cross were 2a:h (8 tetrads), la:2a:1 H (24 tetrads) , 2a:2H (9 tetrads), 2a:1 a:1 H ( 1 tetrad), and la:1 a:2H (3 tetrads). Thirty-six of the 45 tetrads contained a-associated homothallic segregants while 4 of the 45 contained a-associated homothallic spore cultures. In summary, tests of the tetrads 157 and 1009 plus the original observations on homothallic distribution patterns (Table 2) served to identify the presence of two independent allele-specific homothallic factors (H, and H,) in the hetero- thallic parents Y123 and Yl4respectively. Characteristics of homothallic cultures: As mentioned earlier, homothallism in S. cereuisiae and related species arises from a change in the expression of the mating locus from one allele to the other. If H, and H, produce homothallism in S. lactis through an analogous effect, one would expect the homothallic strains to consist of a mixture of competent a and cells. To test this possibility, cells from each of the four homothallic genotypes, &H,, aH,H,, aH,h, and aH,H,, were challenged with a and testers as described under MATERIALS AND METHODS. The number of zygotes formed in the presence of excess a or a cells was an indi- cation of the number of competent or a cells present in the homothallic culture. The results are shmin Table 4. Each representative strain contained both a and 01 cells. A comparison of the ratio of competent a:a cells shows that in the two a-bearing classes, aHah, and aH,H,, cells of a mating type predominated. In the two a-bearing series, ah,H, and aH,H,, on the other hand, there was an excess of cells of 0: mating type. Thus homothallic lines containing the a allele could be distinguished from those with the allele by differences in the ratio of a:acells present in the population. As can be seen in Table 4, a marked difference in the a:a ratio exists between the two a-bearing homothallic lines aH,H, and aH,h,. This difference can be used as a basis for distinguishing between these two genotypes. On the other hand, no significant differences were observed in the ratio of a:acells between the two a-bearing homothallic genotypes &Ha and aH,H, and, therefore, these two genotypes can not be distinguished from one another by this procedure. It can be seen from Table 4 that the 01 homothallic members showed a lower percent homothallism than the a type. This was observed consistently, but at the present time no explanation for this behavior is evident. Perhaps aH, is a more HOMOTHALLISM IN YEAST 733

TABLE 4

Mating response of representatiue clones from the four homothallic classes aH,h,, aH,Ha, ahaH,, and aH,H,

Percent Percent competent Percent competent Ratio competent Geriotvpe hnmothall~sm' a cells a cells n:a cells aHahe 155C 4.4 7.1 2.0 0.3 338C 0.6 7.2 0.2 0.03 UHaH, 1009B 12 4.7 4.2 0.9 155B 11 7.4 6.3 0.9 157A 6 3.5 2.5 0.7 157D 20 15 12 0.8 3381) 7 6.1 3.9 0.6 4,Ha 729B 1.5 2.7 8 3 155A 0.6 0.6 8 13 338D 0.6 0.2 7.2 36 599B 0.5 0.4 5.9 14 ffH,H, 31 1C 1.9 1.1 3.6 3

11st.d rlclerinining percent homothallism and percent competent a and 1y cells are described in MATERIALS and stable gene combination than aH,. In all of these measurements, the percent homothallism expressed by the clones was stable and reproducible 'with a maxi- mum error of 4 to 5 %. Expression of ah,H, homothallism: The preceding study indicated that each of the four classes of homothallic cultures (aHaHa,dah,, aHaH,, and (YhaH,) con- tained mixtures of a and (Y cells. This suggested that Ha and Ha caused a shift in expression of the mating type allele from a to (Y or vice versa. The initial segre- gation patterns for homothallism had led to the assumption, furthermore, that the effects of Ha and Hawere directed in a specific manner against the a and a alleles respectively. If this were the case, one should be able to predict the mating re- sponse of segregants arising from the mating of sister cells in a homothallic cul- ture. If Ha, for example, converted the (Y allele in cells of an (YhaHa clone to the a allele, these newly derived a cells should be (1) unaffected by the Ha gene, (2) heterothallic, and (3)of a mating type. Consequently segregants derived from a mating of an &,Ha cell with such a newly derived a sister cell should display a 2:2 segregation for homothallism: a mating types. To test these predictions, the mating response of &,Ha homothallic clone, 729B, as well as that of spore isolates obtained from one of its tetrads, 1216, was studied. The results are shown in Table 5. Most of the competent cells in the homothallic parent 729B were of (Y mating type. Two of the segregants (1216A and 1216C) derived from a self-sporulated ascus of 729B were homothallic and responded to tester cells in a manner analogous to the homothallic parent 729B. On the other hand, the other two segregants, 1216B and 1216D, formed zygotes with the (Y 734 A. HERMAN AND H. ROMAN TABLE 5 Percent homothallism and mating response of homothallic parent ah,H, (7298) and progeny derived from its self-sporulated tetrad 1216

Percent competent Ratio Percent competent Culture number Genotype homothallism a cells a cells a:a cells 729B ahaH, 1.5 8 2.7 3 1216A ahaHm 3.9 11 3 3 1216B ahaHa 0 0 25 0 1216C “haHa 6.3 4 2.5 1.6 1216D ah,Ha 0 0 16 0 tester strain only, i.e., they behaved as stable heterothallic cells of a mating type. Since these a segregants were derived from a self-sporulated &Ha zygote, their complete genotype can be expressed as ahaH,. Further, the newly derived a cells showed a normal response to the effects of the Ha gene. For example, when one of the a segregants, 1216B, was mated with the heterothallic parent &,ha (Y123), typical aH, recombinant homothallic progeny were obtained. The complete seg- regations for this cross ah&?@X 01H,h, (1216B x Y123) included 201:2a (4 tet- rads), h:la: 1 H (2 tetrads), 201:2H ( 1 tetrad), and 101:la: 2H (1 tetrad). From these results it is concluded that Harenders 01 strains homothallic by changing the expressed mating type of some cells within the culture from 01 to a. Expression of aHahahomothllism: Since clones of aHahand aH,Ha genotypes consisted of a mixture of a and 01 cells (Table 4), it was assumed that Ha, in a manner analogous to the action of the Ha gene, effects a shift of the mating allele from a to 01. If this were true, one would expect a 2:2 segregation of homothallism: 01 mating type in tetrads derived from self-sporulated aH,ha cultures. When self-sporulated tetrads obtained from two independent aHaha clones (155C and 338C) were examined, the results shown in Table 6 were obtained.

TABLE 6

Comparison of percent homothallism and mating response of two homothallic aHah, clones (155C and 338C) and their progeny

Percent competent Ratio Percent competent Culture number Genotype homothallism a cells a cells a:a cells 155C aHaha 4.4 2.0 7.1 3.5 251A aH,ha 3.2 5.0 2.4 0.5 251B aH,ha 5.9 2.8 10.0 3.6 251C aHaha 5.5 2.7 7.8 2.9 251D &Haha 3.0 5.3 2.6 0.5 338C aH,ha 2.4 0.4 7.0 17.5 953A nH,ha 1.5 4.3 0.4 0.1 953B aH,ha 3.0 3.9 4.1 1.1 953c aH,ha 4.2 2.6 2.8 1.1 953D aH,ha 0.7 5.8 1.2 0.2 HOMOTHALLISM IN YEAST 735 Tetrad 251 was derived from homothallic culture 155C and 953 from 338C. Con- trary to expectation, it can be seen that instead of the predicted 2:2 ratio of homo- thallism to mating type, all four segregants from each tetrad were homothallic. This discrepancy will be considered in more detail later. First let us compare the relative proportions of competent a:a cells present in each segregant. It should be noted that two of the segregants in 251 and 953, although homothallic, each ex- hibited a clear-cut excess of a-competent cells, whereas the parental cultures 155C and 338C both had an excess of a-competent cells. This suggests that H, has induced a change from a to a but that this change is unstable in these tetrads. Characteristics of newly deriued aH,h, cells: If, as the evidence indicates. H, converts the mating potentialities of a fraction of the cell population from the a to the form, then why do these H,-induced a cells remain homothallic? It was possible that in the H, mediated conversion of a to a something analogous to phenotypic lag occurred, i.e., after the shift from a to 01, dilution of some remain- ing type component was necessary before the H,-induced a cells became phenotypically heterothallic and of a mating type. To test this suggestion, the homothallic potentialities and mating response of the H,-induced strain 251A were examined at monthly intervals for a period of 3 months. Between examinations, the culture was maintained on stock media and transferred at weekly intervals. Consequently, numerous generations of vege- tative growth occurred between consecutive comparisons. The percent homothal- lism varied from one test to another by 0.8% while the ratio of competent a:a cells remained at approximately two. Thus, no change in the homothallic poten- tialities or mating response occurred over this time interval. Therefore, dilution of some “a” mating type component via vegetative growth seemed unlikely. The two following alternative possibilities were considered: (1) The instability might reflect some property unique to the aH,h, parents from which the a strains were derived rather than that of the new a allele itself; or (2) the instability resulted from characteristics specific to the newly derived a cell itself. The first alternative was considered by examining the mating response of the ct progeny derived from a cross of a homothallic aH,h, (155C) strain with a known heterothallic aH,h, (Y123) line. Since the parents were homozygous for the ha allele, all of the a-bearing progeny should be heterothallic. However, if the aH,h, ( 155C) parent possessed some additional unrelated property which ren- dered the a allele unstable, then some, if not all, of the a-bearing progeny should become homothallic. This was not observed experimentally. In each of six tetrads from the cross aH,h, X aH,h, (155C X Y123) all of the a-bearing progeny were heterothallic. The segregation ratios in each case were 2a:2H. Thus, one could ~liininatethe first alternative. ‘To test the second alternative, that the observed instability in mating type was due to some quality unique to the newly derived a cell, the homothallic behavior of segregants derived from the self-sporulated unstable aH,h, culture 251A (from 155C, Table 6) was examined for evidence that the instability was trans- missible. In culture 251A the ratio of zygotes:100 vegetative cells averaged three, and there were approximately twice as many competent a cells as a cells within 736 A. HERMAN AND H. ROMAN

TABLE 7

Comparison of percent homothallism and mating response of homothallic parent cuH,h, (25IA) and progeny derived from two self-sporulated tetrads 1254 and 1256

Percent competent Ratio Percent competent Culture number Genotype honiothallism N cells a cells a:a cells 251A 3.2 5.0 2.4 2.1 1254A 0.6 4.9 0.3 16.3 1254B 0.5 13.0 0.6 21.6 1254.C 6.0 3.0 7.0 0.5 1254D 3.0 1.5 7.0 0.2 1256A 6.4 3.0 11.0 0.3 1256B 6.0 3.0 13.0 0.2 1256C 0.6 11.9 0.3 39.6 l256D 1.4 6.7 1.o 6.7 the population. When asci from 251A were dissected and the mating potentials of the segregants were compared to those of 251A, one observed marked differ- ences in the mating response of some of the segregants. The results from two such self-sporulated tetrads, 1254 and 1256, are shown in Table 7. In 1254, the homo- thallic potentialities of two of the segregants, 1254A and 1254B, displayed a marked decrease fivefold to sixfold from that of parent 251A. Further, when these isolates were challenged with a and testers, they exhibited a 16-fold to 22-fold excess of competent CY cells compared to a cells. Thus, following one meiotic passage the unstable a segregants of presumed aHahNgenotype had undergone a fivefold reduction in their homothallic potentialities while the ratio of competent a:a cells had increased by a factor of 8 to 11 over the aHahaparent 25 1A. In the second tetrad, 1256, the a segregants, 1256C and 1256D, displayed a similar pattern of behavior (Table 7). Each isolate showed a reduced percentage of homothallism and an increased ratio of competent m:a cells. However, the degree of stabilization exhibited by these two CY sister segregants was noticeably different. In 1256C the percent homothallism had been reduced by a factor of five (from 3% in parent 251A to 0.6% in the segregant) while the ratio of com- petent a:a had risen 20-fold (from 2% in parent 251A to 39.6 in 1256C). In isolate 1256D, on the other hand, the difference between parent and segregants was not as marked. The percent homothallism had been reduced by a factor of two and the ratio of competent a:a had shown a threefold increase only. These results indicated that, whatever the nature of this meiotically induced stabiliza- tion, it was not directed with equal intensity against the two CY sister segregants. The percent homothallism continued to decrease following additional meioses. For example, in tetrad 1290, derived from a self-sporulated zygote of 1256C, the a segregants 1290A and 1290B showed a further reduction in homothallic be- havior (0.2% compared to 0.6% in parent 1'256C) while the ratio of competent ~:acells remained high (20- to 22-fold excess) (Table 8). Thus, following three generations of self-sporulation the percent homothallism in Ha-induced N segre- HOMOTHALLISM IN YEAST 737

TABLE 8

Comparison of the percent homothallism and mating response of homothallic parent aHah, (1256C) and its progeny 1290

Percent competent Ratio Percent competent (:u1 lure nnnilipr tienotvne homothallism a cells a cells a:a cells 1256C uHaha 0.6 11.9 0.3 39.6 1290A aH,ha 0.2 4.0 0.2 20 1290B aH,ha 0.2 4.0 0.16 22 1290C aH,h, 11.0 1.5 7.0 0.2 1290D aHah, 6.0 1.2 4.0 0.3 gants decreased from 3% in culture 251A through 0.6% in culture 1256C to 0.2% in strains 1290A and 1290B. This tendency of homothallic aH,h, segregants to progress toward heterothal- lism occurred in tetrads arising from homothallic x heterothallic crosses also. As shown in Table 9, each of the a! segregants arising from the mating of homo- thallic &,ha x heterothallic ahaH, cells (251A x Y14) was more strongly heterothallic than the parent 251A. Further, and as was observed above in the self-sporulated a! segregants, the stabilizing effects exerted on sister a! segregants were not identical and did not appear to follow any pattern. However, disregard- ing these inequalities in effect, it is evident that the reduction in the homothallic tendencies of aHaha cultures depends, apparently, upon some basic mechanism associated with or following meiosis. This continued decrease in homothallic behavior associated with sequential mating and sporulation suggests that, if these processes were continued long enough, one might eventually obtain completely stabilized aH,h, heterothallic strains of a! mating type such as the a! parent Y123 and implies that this may have been the mechanism through which this parent was formed. Possible explanations for the homothallism associated with newly derived H,-induced a! cells will be considered in the DISCUSSION.

TABLE 9

Comparison of the percent homothallism and mating response of homothallic parent aH,h, (251A) and homothallic CY segregants from the mating aHah, x ah,H, (251A x Y14)

~ ~~ ~ Percent competent Ratio Percent competent Ciil ture number Genotype homothallism a cells a cells ,:a cells 251A aHah, 3.2 5.0 2.4 2.1 1255C aHah, 0.4 8.0 0.1 80 1255D aHah, 1.8 4.0 0.4 10 2257A ffHaha 0.1 7.0 70 1257D aHah, 1.1 4.0 0.7 5.7 1258C aHaha 0.9 9.5 0.5 19 125813 offaha 3.1 2.9 0.3 10 738 A. HERMAN AND H. ROMAN

DISCUSSION Evidence from this study indicates that H, and H, produce homothallism by shifting the expressed mating type to the alternate form in a certain portion of the cell population. Except for the allele specificity of H, and H,, their mode of action parallels that of the D gene (OESER1962; HAWTHORNE1963b) quite closely. Saccharomyces lactis does not hybridize with S. cerevisiae and related strains; therefore, one cannot test the homology of the two systems. The nature of the mechanism through which H, and H, produce their effects remains a point of speculation. A possible explanation is that H, and H, produce the shift in mating type by increasing the rate of mutation at the mating locus from one allele to the other. This would, however, require specific mutagenic effects by products from Ha and H, and such effects are as yet unknown. Are H, and H, acting as regulatory genes (JACOBand MONOD1961)? If, as in other yeasts, the mating locus is complex (LEUPOLD1958 and HAWTHORNE 1963a) and consists of a- and a-determining regions one might postulate that H, and H, produce specific repressors that determine cwhich region of the complex locus will be expressed. Such a model proves inadequate, however, since it does not explain the heterothallic nature of the ah,h, and ahh, recombinants which were obtained (Table 2) nor does it account for the stability of the “mutation” once it is induced. Thus no satisfactory explanation for the action of H, and Ha is evident at this time. Also still obscure is the significance of the paradoxical observation relating to the homothallic behavior of Nu-induced a cells. Contrary to expectation, these

H,-induced (Y cells were homothallic, and the degree of homothallism remained constant during an extended period of vegetative propagation; however, the homothallic potentialities of these H,-induced cells were not transmitted intact to their a-bearing progeny as evidenced by the marked decline in the percent homothallism expressed by the segregants. This instability in mating type did not reflect a generalized effect of the aH,h, genome on all a alleles as shown by the fact that normal stable heterothallic segregants of (Y mating type were con- sistently derived from the mating of aH,h, cells with known heterothallic cells of aH,h, genotype. Thus, the observed instability of the newly derived H,-induced a cultures apparently reflected an interaction between the aH,h, parental cells and the newly derived a clones. These observations suggest (1) that the H,-induced 01 cells retained the aH,h, parental capacity to form a mating component even though they lacked the neces- sary genetic information and (2) that this capacity was propagated in some auton- omous manner during vegetative growth but decreased during sporulation. This capacity could reflect either a delay in the accumulation or distribution of a prod- uct determining a mating type or a reduction in the ability of the newly derived a cells to destroy information carried over from the aH,h, parent for determining synthesis of a mating type product. One region of the cell which may serve as the vehicle for transmission and expression of the homothallic potentialities observed in H,-induced strains is HOMOTHALLISM IN YEAST 739 the cell wall. It has been demonstrated that reactions between specific cell wall components implements mating between some heterothallic yeasts (BROCK1959). EDDYand WILLIAMSON( 1959) have suggested that during vegetative propaga- tion yeast cell wall structure is copied from existing cell wall. A continuous flow of nuclear information directing cell wall synthesis may not be necessary. If specificity for a mating type resides in some cell wall component, it is easy to see how, once the a-forming capacity of &Huh,cells was expressed in cell wall structures, its continued synthesis could remain constant during extended periods of vegetative growth. Since it is known that yeast spores possessing normal cell walls are formed from diploid protoplasts devoid of cell walls (EDDYand WILLIAMSON1959), it is apparent that spore wall synthesis must be nuclear controlled rather than initiated from vegetative cell wall primer. The switch from vegetative to nuclear control at this time would, of course, permit full expression of cell wall determinants quite different from those expressed in the original diploid. In the case of the Ha- induced (Y homothallic cells, any preceding events leading to the inactivation of the nonnuclear a-determining information could be expressed at this time as a reduction in the homothallic potentialities of clones derived from the a spores. Finally, one further implication which may be derived from these experiments should be considered briefly. Ha and H, cause the mating locus to shift from one allele to the other. However, this shift in mating type reflects only one aspect of the overall process. As was indicated earlier, the percent homothallism expressed by representatives of the four homothallic classes, aH,H,, aHaha, aH,H, and ah,,H, was reproducible to within a few percent. This constancy in percent homothallism would not be expected if the H,- and Ha-directed shift in mating type had occurred at random and over a period of time in a population of vegeta- tively growing cells. Under such circumstances, one would have observed an alteration in the percent homothallism due to changes in the number of derived U or a cclls which would accumulate in the population. On the contrary, the observed constancy in percent homothallism implies that the H,- and H,-directed shift did not occur during vegetative growth, but appeared after the cells had been transferred to ME only. Thus, it is evident that another independent system of controls must determine when H, and H, will exert their effect. Since nothing is known about the nature of the physiological changes which occur in S. lactis during growth on ME, speculation concerning the nature of the controls which allow Nuand H, to become active during growth on this medium would be fruit- less at this time. The senior author wishes to thank PROFESSORHARLYN 0. HALVORSONin whose laboratory this work was initiated and DR.DONALD C. HAWTHORNEfor his interest and suggestions.

SUMMARY Two unlinked independent genes designated H, and H, produce homothallism in strains of S. lactis. These factors exert their effects by changing the mating type to the alternate form in a small proportion of a cell population. Both H, and H, are allele specific. H, changes (Y mating type to a while H, changes a to a. 740 A. HERMAN AND H. ROMAN

LITERATURE CITED

AHMAD,M., 1952 Single-spore cultures of heterothallic Saccharomyces cereuisiae which mate with both tester strains. Nature 170: 546-547. BROCK,T. D., 1959 Mating reactions in Hansenula wingei. J. Bacterid. 78:59-68. EDDY,A. A., and D. H. WILLIAMSON,1959 Formation of aberrant cell walls and of spores by the growing yeast protoplast. Nature 183: 1101-1 104. H*WTHORNE,D. C., 1963a A deletion in yeast and its bearing on the structure of the mating type locus. Genetics 48: 1727-1729. - 1963b Directed mutation of the mating type allele as an explanation of homothallism in yeast. (Abstr.) Proc. 11th Intern. Congr. Genet. 1: 34-35. HERMAN,A., and H. 0. HALVORSON,1963 Identification of the structural gene for p-glucosidase in Saccharomyces lactis. J. Bacteriol. 85 : 895-900. J~COB,F., and J. MONOD,1961 On the regulation of gene activity. Cold Spring Harbor Symp. Quant. Biol. 26: 193-211. LEUPOLD,U., 1958 Studies on recombination in Schizosaccharomyces pombe. Cold Spring Harbor Symp. Quant. Biol. 23: 161-1 70. OESER,H., 1962 Genetische Untersuchungen iiber das Paarungstypverhalten bei Saccharomyces und die Maltose-Gene einiger untergariger Bierhefen. Arch. Mikrobiol. 44: 47-74. PERKINS,D. D., 1953 The detection of linkage in tetrad analysis. Genetics 38: 187-197. TAKAHASHI,T., 1958 Complementary genes controlling homothallism in Saccharomyces. Genetics 43: 705-714. TAKAHASHI, T., and Y. IKEDA,1959 Bisexual mating reaction in Saccharomyces cheualieri. Genetics 44: 376-382. TAKAHASHI,T., H. SAITO,and Y. IKEDA,1958 Heterothallic behavior of a homothallic strain in Saccharomyces yeast. Genetics 43 : 249-260. WICKERHAM,L. J., 1951 Taxonomy of yeast. U. S. Dept. Agr. Tech. Bull. No. 1029; Washing- ton, D.C. WINGE,0., and C. ROBERTS,1949 A gene for diploidization in yeast. Compt. Rend. Trav. Lab. Carlsberg. 24: 341-3445.